The sponsor is requested to explain the rationale behind the exclusion from Study A5481008 of patients with recent or active suicidal ideation or behaviour. In particular, the sponsor is requested to provide case details where palbociclib might have been implicated in patients committing suicide or becoming suicidal that is, while taking or after recently stopping palbociclib. And further to the results in Study A5481023, where 4 patients were reported to have a psychosis, depression or suicide attempts, the sponsor is requested to discuss the potential role of palbociclib in these events.
The exclusion of patients with recent or active suicidal ideation or behaviour from the A5481008 study protocol is consistent with standard protocol language included in all Pfizer protocol templates. Current Pfizer template language provides the rationale for excluding patients with mental health related diagnosis as follows:
‘acute or chronic medical or psychiatric conditions including recent (within the past year) or active suicidal ideation or behaviour or laboratory abnormality that may increase the risk associated with study participation or investigational product administration or may interfere with the interpretation of study results and, in the judgment of the Investigator, would make the subject inappropriate for entry into the study’.
No cases of suicide attempts are reported in either Study A5481003 or Study A5481008 studies, while in Study A5481023, as this question indicated, 4 patients (3 in the palbociclib plus fulvestrant treatment group and 1 in the placebo plus fulvestrant treatment group) were reported to have serious (Grade 3 or 4) psychiatric disorders adverse events (that is, Psychotic disorder, Depression or Suicide attempts) as follows (Table 188.8.131.52.1 and 184.108.40.206.2 of A5481023 90 Day Safety Update):
Palbociclib plus Fulvestrant Treatment Group
2 patients (0.6%) were reported to have experienced a suicidal attempt: Grade 3 (Patient 11031005) and Grade 4 Suicide attempt (Patient 11481004) respectively.
1 patient (0.3%) was reported to have one episode of Grade 4 Psychotic disorder (Patient 10731014)
1 (0.3%) patient was reported to have experienced Grade 3 Depression (Patient 11661005)
Additional details on these 4 adverse events are provided in Table 1.
The 4 above mentioned adverse events were associated with hospitalization and 1 adverse event (attempted suicide of Patient 11481004) led to permanent discontinuation of palbociclib. Three out of these 4 adverse events occurred while the patient was on study treatment, whereas Patient 11031005 attempted suicide 10 days after study treatment was permanently discontinued due to disease progression.
Of note, these events were all considered to be unrelated to study drug by the Investigators. The reported 4 psychiatric adverse events occurred in patients with concurrent depression, which in itself may lead to suicidal ideation or behaviour. In addition one patient (Patient [information redacted]) also had a relevant medical history for psychiatric disorders, had experienced previous episodes of psychosis and was under antipsychotic medical treatment at study entry. The psychotic episode for this patient that occurred on study was attributed by the Investigator to the permanent withdrawal of a concomitant antidepressive drug (quetiapine fumarate) three days before the first dose (Cycle 1 Day 1). The reported adverse event occurred approximately 3 months later (Cycle 4 of study treatment) and fully recovered with quetiapine fumarate re-introduction in patient’s therapy.
Patients enrolled in Study A5481023 have experienced the challenges of disease progression and the limitations of existing therapeutic options available to them. The underlying advanced malignancy and associated psychological impact on the patient must also be considered as factors that may favor the development of or the worsening of pre-existing depression. In Study A5481023 there were alternative explanations for each of the 4 cases and these events were assessed by the Investigators as being unrelated to study drug.
The sponsor believes that there is insufficient evidence to consider palbociclib to be causally related to the development of psychiatric disorders.
The inclusion of the following exclusion criterion in Study 1023 supports the sponsor’s response that this criterion was not specific to Study 1008. The reasoning for including this specification in protocols may stem from its inclusion in trials of medications for depression.
14. Other severe acute or chronic medical or psychiatric condition, including recent or active suicidal ideation or behaviour, or laboratory abnormality that might have increased the risk associated with study participation or investigational product administration or might have interfered with the interpretation of study results and, in the judgement of the investigator, would have made the patient inappropriate for entry into this study.
The wording of the same criterion is less specific in the CSR for Study 1003 however it is noted that this may be a difference between Phase I/II and Phase III trial protocol templates.
Other severe acute or chronic medical or psychiatric condition or laboratory abnormality that may have increased the risk associated with study participation or investigational product administration or may have interfered with the interpretation of study results and, in the judgment of the investigator, would have made the patient inappropriate for entry into this study.
Given the circumstances they are in, I would be surprised if the rate of observed events (1.2%) was higher than the background rates of depression and suicide ideation, given their circumstances.
For points of comparison: rates of completed suicide in breast cancer survivors in the USA and Scandinavia have been reported to be between 0.07% and 0.99%.29 Rates of major and minor depression in a sample of 200 mostly Australian metastatic breast cancer patients were reported to be 6.5% and 24.5% respectively.30
Table 42: Rates of completed suicide in various breast cancer registry studies (Source data: Guth et al.29)
Source registry and timeframe
Schairer et al31: 1953-2001
US SEER Program: 1973–2001
Basel Breast Cancer Database: 1990-2006
Although there was an imbalance seen between active and comparator arms in Study 1023 with regard to psychiatric disorders, statistically this is likely to have been by chance, as the absolute numbers are small. Whether this is a pattern that has been noted in the safety data of confirmatory trial 1008 is not clear from the s31 responses, and should be checked on review of that study.
The sponsor’s response is accepted.
Safety clinical question 3
The sponsor is requested to provide with a future provision of the CSR for Study 1008, an integrated safety summary for the 2 studies 1003 and 1008, and an updated PI to reflect these data.
The sponsor holds the evidence to which this Section 31 request refers. The sponsor has therefore appended the Clinical Study Report for Study A5481008 in its entirety as the source document for the Section 31 request for information questions relating to Study A5481008. The sponsor believes the provision of Study A5481008 CSR and the corresponding Summary of Clinical Safety (SCS) addresses the request for an integrated safety summary for Studies A5481003 and A5481008, for the following reasons:
Study A5481008 was a large randomised, double –blind phase 3 study designed to be a confirmatory trial of Study A5481003. Study A5481008 was conducted in a similar population of postmenopausal women with ER-positive HER2-negative advanced breast cancer as Study A5481003 but with a much larger sample size than Study A5481003 (N=666 for Study A5481008 versus N=165 for Study A5481003, with 444 patients in Study A5481008 and 83 patients in Study A5481003 treated with palbociclib, respectively).
The results described in the Study A5481008 CSR and SCS clearly confirm the clinical benefit of palbociclib and letrozole with a statistically significant, robust and clinically meaningful 10-month improvement in median PFS and also demonstrate that the safety profile of the combination of palbociclib plus letrozole was very similar in both studies.
The sponsor has also conducted an analysis of the frequency and severity of treatment-emergent adverse events in a pooled dataset of Studies A5481003 and A5481008. The frequency and severity of treatment-emergent adverse events were generally comparable to those observed in Studies A5481003 and A5481008 when analysed individually.
To support the above considerations, the following pooled safety tables are provided:
Summary of treatment-emergent adverse events by MedDRA System Organ Class, Preferred Term (including clusters) and maximum CTCAE grade ( all causality) – Tables 295.1, 295.2
Summary of treatment-emergent adverse events associated with permanent discontinuation, dose reduction or temporary discontinuation (all causality) – Tables 279.1 to 279.6
Summary of time to first onset of neutropenia and summary of duration of neutropenia (based on laboratory test data) – Tables 322.1 and 322.2
For neutropenia, the most common palbociclib-related adverse event, it is important to note the median time from first dose to first episode onset is exactly the same (15 days) in the integrated analysis (pooled data from Studies A5481003 and A5481008) as in Study A5481008 alone results (Table 322.1 and Table 51 of SCS). The median duration of any grade neutropenia is also similar between the pooled dataset of Studies A5481003 and A5481008 and Study A5481008 alone analysis (316 versus 321 days respectively: Tables 322.2 and Table 52 of SCS).
The sponsor believes that the information provided in the CSR for Study A5481008 confirms the information already provided in the registration application. Thus, an integrated safety summary of Studies A5481003 and A5481008 would not be expected to provide relevant additional information on the safety of the palbociclib plus letrozole combination.
It is noted, that an integrated safety summary for A5481003 and A5481008 has not been requested by any other major regulator, including the EMA for which a full approval was granted based on the same data set provided in the initial registration application to the TGA. The EMA also received the 90 Day Safety Update from the sNDA for Study A5481023. An updated PI presenting the combined adverse drug reactions for Studies A5481003 and A5481008 for the treatment and comparator arms is provided as part of the Section 31Response for this application.
The tables referred to by the sponsor have not been appended to this CER due to their length. Bone marrow suppression, primarily manifested by Grade 3-4 neutropenia, is the predominant feature of the AE profile for the active compared to the placebo arm.
Study 1003 is of lower phase than Study 1008 and had a number of issues around imbalanced randomisation (see evaluator comments following Efficacy question 3). Study 1008 is a much more robust demonstration of both efficacy and safety, and the safety data submitted with the full CSR will be reviewed in the context of the study.
The sponsor’s response is accepted.
Safety clinical question 4
For Study A5481023, the remaining events accounting for the 3.5% of Grade 3 or 4 TEAEs for the palbociclib and fulvestrant arm and 1.7% in the comparator arm (90-day safety update) are not presented. The sponsor is requested to provide this information.
The sponsor would like to clarify that in the in-text Table 7 (Summary of All-Causality, Treatment-Emergent Adverse Events (All Cycles) Experienced by at Least 10% of Patients in Either Treatment Arm of Study A5481023 as of 31 July 2015 by MedDRA PT and Maximum Severity Grade Sorted by Decreasing Frequency (All Severity Grades) for Patients Receiving Palbociclib Plus Fulvestrant – All Treated Patients) of the 90 Day Safety Update document all the TEAEs experienced by at least 10% of patients in either treatment arm were reported as indicated in the table title.
In the footnote of Table 7 the number of the source table from which Table 7 was derived is provided (Table 220.127.116.11.3). This table was included together with all other source tables at the time the Study A5481023 90 Day Safety Update (data cutoff 31 July 2015) was submitted to the TGA. Table 18.104.22.168.3 includes all the TEAEs of any frequency that were reported by the defined data cutoff date by patients in the 2 treatment arms.
This table is linked directly to this response for convenience.
Table 22.214.171.124.3 contains a listing of all TEAEs in both arms in Study 1023.
The Grade 3 and 4 events have been reviewed (see Table 43) and do not reveal any new safety signals. The size of the study is not large enough to draw significant comparisons between arms where there are isolated cases. Additionally, multiple preferred terms (for example, fracture femur fracture, humerus fracture and road traffic accident) may have been reported for one case.
Neutropenia and Grade 3 and 4 infections were reported more frequently in the active arm. Additionally, there were 7 reports of ‘breast mass’ in the active arm: an AE term which is clearly related to underlying disease. It is very possible that a reporting bias is present, in that the presence of neutropenia could have resulted in partial unmasking of treatment. The other notable imbalance is in the SOC psychotic disorders. Although, like road traffic accident related fractures and a pathological fracture this is likely a reflection of chance rather than ascribable risk, this aspect of safety should be given particular consideration in review of Study 1008 data (see section 13).
Table 43: Grade 3 and 4 TEAEs in Study 1023, by treatment and System Organ Class (SOC)
An update infection rate/febrile neutropenic rate has not been presented in this safety update for Study 1023 and this is needed, to ensure the figure quoted in the PI is accurate.
Please refer to the sponsor’s response to PI Question 23 and PI Question 26 which duplicate this query Safety Question 5.
The sponsor’s response is accepted.
Safety clinical question 6
Bone marrow suppression occurs resulting in leukopenia and neutropenia. The sponsor is requested to provide the following information, and where cases have occurred, provide the details. What were the rates of opportunistic infections reported for each of the arms for Study 1003, 1023 and 1008?
Were any cases of Hepatitis B reactivation identified?
Has PML ever been reported in the palbociclib development program?
Please provide details of any cases reported.
Response to Query 6a:
In order to evaluate the frequency of opportunistic infections in Studies A5481003, A5481008, and A5481023, a search was performed for the following reported Preferred Terms (PTs) through January 02, 2015 (A5481003), 26 February 2016 (A5481008), and July 31, 2015 (A5481023):
A 72 year old obese (99 kg) female ([information redacted]) with non-insulin dependent diabetes mellitus, treated with palbociclib plus letrozole in Study A5481003, developed Grade 2 Vulvovaginal candidiasis on Study Day 161. The event resolved on Study Day 168. No action was taken with study medications.
A 70 year old obese (107 kg) female ([information redacted]) treated with palbociclib plus letrozole in Study A5481008 developed Grade 1 Skin candida on Study Day 44. The event resolved on Study Day 59. No action was taken with study medications.
A 66 year old obese (121 kg) female ([information redacted]) treated with palbociclib plus letrozole in Study A5481008 developed Grade 1 Staphylococcal skin infection on Study Day 157. The event resolved on Study Day 168. No action was taken with study medications.
A 70 year old (51 kg) female ([information redacted]) treated with palbociclib plus letrozole in Study A5481008 developed Grade 2 Staphylococcal infection (‘Staph infection left arm’) on Study Day 220. The event resolved on Study Day 231. No action was taken with study medications.
A 67 year old (61 kg) female ([information redacted]) with a history of hyperlipidaemia, treated with palbociclib plus letrozole in Study A5481008 for recurrent metastatic (sacrum, spine, ribs, bilateral iliac, sternum) breast cancer, developed Grade 2 Clostridium difficile infection on Study Day 90. She was treated with metronidazole and the event resolved 10 days later. Of note, the patient was treated with palliative radiation to the pelvis on Study Days 19 through 29 and was treated with methadone and dexamethasone for bone pain. She was diagnosed with radiation enteritis (PT = Gastroenteritis radiation, Case number [information redacted]) on Study Day 38 and was hospitalized. The event resolved 8 days later. The patient finally developed Grade 3 pneumonitis on Study Day 76 at which time study medications were permanently discontinued. Thus, the patient had not received palbociclib for 2 weeks at the time she developed Clostridium difficile infection. The event was not causally attributed to study medications by the investigator.
A 54 year old (77 kg) female ([information redacted]) with no relevant medical history treated with palbociclib plus letrozole in Study A5481008 for metastatic breast cancer, developed non-serious Grade 3 neutropenia on Study Day 16 and palbociclib was temporarily held. On Study Day 30, palliative radiotherapy was initiated, lasting through Study Day 45. Two days later, on Study Day 47, the patient developed nausea, vomiting, as well as deterioration of her general status, and was hospitalized ([information redacted]) and diagnosed with Clostridium difficile infection. At this time her laboratory values were as follows: haemoglobin: 10.5 g/dL (11.5 to 15g/dL); white blood cell count: 3.5/nL (3.5 to 10/nL). She was discharged 4 days later and considered recovered 8 days after diagnosis, upon which study medications were restarted. The event was not causally attributed to study medications by the investigator.
A 39 year old (65 kg) female ([information redacted]) treated with palbociclib plus fulvestrant in Study A5481023 developed Grade 2 Vulvovaginal candidiasis on Study Day 92. The event resolved on Study Day 294. No action was taken with study medications.
B. Patients treated in comparator arms [N=471]
A 63 year old (57 kg) female ([information redacted]) treated with placebo plus letrozole in Study A5481008 developed Grade 2 Staphylococcal infection on Study Day 162. The event resolved on Study Day 197. No action was taken with study medications. A 70 year old (77 kg) female (Patient ID [information redacted]) treated with placebo plus letrozole in Study A5481008 developed Grade 1 Skin candida on Study Day 420. The event has not resolved as of Study Day 504 and is ongoing. No action was taken with study medications.
Summary and Conclusion:
In summary, among 884 patients treated in the palbociclib arms, 3 candida skin infections and 2 staphylococcal skin infections, and 2 Clostridium difficile infections were observed (rate of 0.8%). Among 471 patients treated in comparator arms, 1 Staphylococcal infection and 1 candida skin infection were observed (rate of 0.4%). Of note, the adverse event analysis was not adjusted for the significantly longer study treatment duration in the palbociclib-treated arms compared to the treatment duration in the comparator arms (Study 1003: 420 days vs. 231 days; Study 1008: 603 days vs. 413 days; Study 1023: 330 days vs. 137 days, respectively).
Skin infections, particularly vulvovaginal candidiasis, as well as staphylococcal skin infections, but also clostridia infections (or carriage) may occur in patients without obvious immunocompromise. Diabetes (present in patient [information redacted]) as well as obesity (present in 3/7 palbociclib treated patients), cancer, and recurrent hospitalizations, as well as treatment with antibiotics are risk factors for these infections. Cutaneous and/or systemic candida or staphylococcal infections occur with much higher frequency and severity in immunocompromised patients and are therefore also considered under opportunistic infections. However, they are not exclusively or necessarily classic opportunistic infections, and thus their occasional occurrence at low severity grades in palbociclib treated patients does not justify the conclusion that palbociclib increases patients’ risk of developing opportunistic infections. When diagnosing clostridia infections, it is important to distinguish a positive carrier status from actual infection. No clinical details are provided for Patient ID [information redacted], thus it is not clear what, if any, symptoms were present. Patient [information redacted] did have gastrointestinal symptomatology and required hospitalization. Both patients had preceding radiotherapy, which may have caused or exacerbated gastrointestinal symptoms.
In patient [information redacted] palbociclib had not been administered for 30 days prior to the diagnosis of Clostridium difficile infection, while in patient [information redacted], palbociclib had not been administered for 2 weeks at the time Clostridium difficile infection was diagnosed.
Overall, this analysis does not support the conclusion that palbociclib increases a patient’s risk of developing opportunistic infections.
Response to Query 6b:
In order to evaluate the reporting frequency of reported cases of Hepatitis B reactivation, a search of the clinical database was performed for the following Preferred Terms: Hepatitis B (corresponds to Lower Level Term of Hepatitis B reactivation)
Chronic hepatitis B (corresponds to Lower Level Term of Chronic hepatitis B reactivation)
The search did not reveal any cases. In addition, a search was performed of any post-marketing cases of Hepatitis B or Chronic hepatitis B received on the Safety Database through November 15, 2016. This search also did not identify any cases.
Summary and Conclusion:
Testing for hepatitis B infection was not mandated at Study entry and a history of hepatitis B infection was not an exclusion criterion in any of the pivotal breast cancer trials. Since no cases of hepatitis B reactivation were reported in 884 palbociclib treated patients, or in any of the over 40,000 patients treated outside of company sponsored clinical trials, there is currently no evidence that treatment with palbociclib is causally associated with hepatitis B reactivation.
Response to Query 6c:
In order to evaluate the reporting frequency of Progressive multifocal leukoencephalopathy (PML), a search of the clinical database was performed for the following Preferred Terms: Progressive multifocal leukoencephalopathy. The search did not reveal any cases.
In addition, a search was performed of any post-marketing cases of Progressive multifocal leukoencephalopathy received on the Safety Database through November 15, 2016. This search also did not identify any cases.
Summary and Conclusion:
There is currently no evidence that treatment with palbociclib predisposes to the development of PML.